International Journal of Technology Assessment in Health Care最新文献

Objective: This study examined the application, feasibility, and validity of supervised learning models for text classification in appraisals for rare disease treatments (RDTs) in relation to uncertainty, and analyzed differences between appraisals based on the classification results.

Methods: We analyzed appraisals for RDTs (n = 94) published by the National Institute for Health and Care Excellence (NICE) between January 2011 and May 2023. We used Naïve Bayes, Lasso, and Support Vector Machine models in a binary text classification task (classifying paragraphs as either referencing uncertainty in the evidence base or not). To illustrate the results, we tested hypotheses in relation to the appraisal guidance, advanced therapy medicinal product (ATMP) status, disease area, and age group.

Results: The best performing (Lasso) model achieved 83.6 percent classification accuracy (sensitivity = 74.4 percent, specificity = 92.6 percent). Paragraphs classified as referencing uncertainty were significantly more likely to arise in highly specialized technology (HST) appraisals compared to appraisals from the technology appraisal (TA) guidance (adjusted odds ratio = 1.44, 95 percent CI 1.09, 1.90, p = 0.004). There was no significant association between paragraphs classified as referencing uncertainty and appraisals for ATMPs, non-oncology RDTs, and RDTs indicated for children only or adults and children. These results were robust to the threshold value used for classifying paragraphs but were sensitive to the choice of classification model.

Conclusion: Using supervised learning models for text classification in NICE appraisals for RDTs is feasible, but the results of downstream analyses may be sensitive to the choice of classification model.

Objectives: Lack of evidence regarding safety and effectiveness at market entry is driving the need to consider adopting a lifecycle approach to evaluating medical devices, but it is unclear what lifecycle evaluation means. This research sought to explore the tacit meanings of "lifecycle" and "lifecycle evaluation" as embodied within evaluation models/frameworks used for medical devices.

Methods: Drawing on qualitative evidence synthesis methods and using an inductive approach, novel methods were developed to identify, appraise, analyze, and synthesize lifecycle evaluation models used for medical devices. Data was extracted (including purpose; audience; characterization; outputs; timing; and type of model) from key texts for coding, categorization, and comparison, exploring embodied meaning across four broad perspectives.

Results: Fifty-two models were included in the synthesis. They demonstrated significant heterogeneity of meaning, form, scope, timing, and purpose. The "lifecycle" may represent a single stage, a series of stages, a cycle of innovation, or a system. "Lifecycle evaluation" focuses on the overarching goal of the stakeholder group, and may use a single or repeated evaluation to inform decision-making regarding the adoption of health technologies (Healthcare), resource allocation (Policymaking), investment in new product development or marketing (Trade and Industry), or market regulation (Regulation). The adoption of a lifecycle approach by regulators has resulted in the deferral of evidence generation to the post-market phase.

Conclusions: Using a "lifecycle evaluation" approach to inform reimbursement decision-making must not be allowed to further jeopardize evidence generation and patient safety by accepting inadequate evidence of safety and effectiveness for reimbursement decisions.

Introduction: The adoption of genomic technologies in the context of hospital-based health technology assessment presents multiple practical and organizational challenges.

Objective: This study aimed to assist the Instituto Português de Oncologia de Lisboa Francisco Gentil (IPO Lisboa) decision makers in analyzing which acute myeloid leukemia (AML) genomic panel contracting strategies had the highest value-for-money.

Methods: A tailored, three-step approach was developed, which included: mapping clinical pathways of AML patients, building a multicriteria value model using the MACBETH approach to evaluate each genomic testing contracting strategy, and estimating the cost of each strategy through Monte Carlo simulation modeling. The value-for-money of three contracting strategies - "Standard of care (S1)," "FoundationOne Heme test (S2)," and "New diagnostic test infrastructure (S3)" - was then analyzed through strategy landscape and value-for-money graphs.

Results: Implementing a larger gene panel (S2) and investing in a new diagnostic test infrastructure (S3) were shown to generate extra value, but also to entail extra costs in comparison with the standard of care, with the extra value being explained by making available additional genetic information that enables more personalized treatment and patient monitoring (S2 and S3), access to a broader range of clinical trials (S2), and more complete databases to potentiate research (S3).

Conclusion: The proposed multimethodology provided IPO Lisboa decision makers with comprehensive and insightful information regarding each strategy's value-for-money, enabling an informed discussion on whether to move from the current Strategy S1 to other competing strategies.

Health technology assessment (HTA) programs inform decision making about the value and reimbursement of new and existing health technologies; however, they are under increasing pressure to demonstrate that they are a cost-effective use of finite healthcare resources themselves. The 2023 HTAi Global Policy Forum (GPF) discussed the value and impact of HTA, including how it is assessed and communicated, and how it could be enhanced in the future. This article summarizes the discussions held at the 2023 HTAi GPF, where the challenges and opportunities related to the value and impact of HTA were debated. Core themes and recommendations identified that defining the purpose of value and impact assessment is an essential first step prior to undertaking it, and that it can be done through the use and expansion of existing tools. Further work around aligning HTA programs with underlying societal values is needed to ensure the long-term value and impact of HTA. HTA could also have a role in assessing the efficiency of the wider health system by applying HTA methods or concepts to broader budgetary allocations and organizational aspects of health care. Stakeholders (particularly patients, industry, and clinicians but also payers, wider society, and the media) should ideally be actively engaged when undertaking the value and impact assessment of HTA. More concerted efforts in communicating the role and remit of HTA bodies would also help stakeholders to better understand the value and impact of HTA, which in turn could improve the implementation of HTA recommendations and application to future actions in the lifecycle of technologies.

Objectives: Qualitative methods allow in-depth exploration of patient experiences and can provide context for healthcare decision making. Frameworks for patient-based evidence in health technology assessment (HTA) are expanding; yet, how extensively qualitative methods are currently used is unclear. This review characterized the extent and quality of qualitative data submitted to National Institute for Health and Care Excellence (NICE) and Canadian Agency for Drugs and Technologies in Health (CADTH) for HTA.

Methods: NICE and CADTH submissions from September 2019 to August 2021 were reviewed. Submission characteristics and features of patient-based evidence included within submissions were extracted. The quality of qualitative reporting was assessed using the CASP checklist.

Results: Patient-based evidence was included in 83/107 NICE and 119/124 CADTH submissions. A small proportion described qualitative data collection (NICE=14; CADTH=24) and analysis (NICE=6; CADTH=9) methods. One-to-one interviews were the most common data collection method, and thematic analysis was exclusively used. Thirty-three percent of NICE submissions scored >7 yes responses on CASP, versus 78 percent of CADTH submissions.

Conclusions: Although patient-based evidence was common in the submissions reviewed, only 14/107 NICE and 24/124 CADTH submissions involved formal qualitative data collection. Use of formal analysis methods was even rarer and reporting tended to be brief. At present, there is little guidance about qualitative evidence most likely to be informative and therefore to potentially impact decision making. Ensuring, however, that qualitative data are collected and analyzed in a systematic, rigorous way will maximize their usefulness and ensure that patient voices are clearly heard.

Objectives: Published evidence on health service interventions should inform decision-making in local health services, but primary effectiveness studies and cost-effectiveness analyses are unlikely to reflect contexts other than those in which the evaluations were undertaken. A ten-step framework was developed and applied to use published evidence as the basis for local-level economic evaluations that estimate the expected costs and effects of new service intervention options in specific local contexts.

Methods: Working with a multidisciplinary group of local clinicians, the framework was applied to evaluate intervention options for preventing hospital-acquired hypoglycemia. The framework included: clinical audit and analyses of local health systems data to understand the local context and estimate baseline event rates; pragmatic literature review to identify evidence on relevant intervention options; expert elicitation to adjust published intervention effect estimates to reflect the local context; and modeling to synthesize and calibrate data derived from the disparate data sources.

Results: From forty-seven studies identified in the literature review, the working group selected three interventions for evaluation. The local-level economic evaluation generated estimates of intervention costs and a range of cost, capacity and patient outcome-related consequences, which informed working group recommendations to implement two of the interventions.

Conclusions: The applied framework for modeled local-level economic evaluation was valued by local stakeholders, in particular the structured, formal approach to identifying and interpreting published evidence alongside local data. Key methodological issues included the handling of alternative reported outcomes and the elicitation of the expected intervention effects in the local context.

The Health Technology Assessment (HTA) process aims to optimize health system funding of technologies. In recent years there has been an increase in what is known as Real-World Evidence (RWE) as a complement to clinical trials. The objective of Health Technology Assessment International’s Latin American Policy Forum 2022 was to explore the utility of incorporating RWE into HTA and decision-making processes in the region.

This article is based on a background document, survey, and the deliberative work of the country representatives who participated in the Forum.

There is a growing interest in the use of Real-World Data / Real-World Evidence in HTA processes in Latin America, although currently there are no specific local guidelines for RWE use by HTA agencies. At present, its use is limited to certain areas such as adding context to HTA reports, the evaluation of adverse events, or cost estimation.

Potential future uses of RWE were identified, including the creation of risk-sharing agreements, the assessment of technology performance in routine practice, providing information on outcomes that are not so easily evaluated in clinical trials (e.g., the identification of specific subpopulations or quality of life), and the estimation of input parameters for economic evaluations.

The participants agreed that there are several areas presenting significant potential to expand the application of RWD/RWE and that the development of normative frameworks for its use could be helpful.

Objectives: Telemedicine may improve healthcare access and efficiency if it demands less clinician time than usual care. We sought to describe the degree to which telemedicine trials assess the effect of telemedicine on clinicians' time and to discuss how including the time needed to treat (TNT) in health technology assessment (HTA) could affect the design of telemedicine services and studies.

Methods: We conducted a scoping review by searching clinicaltrials.gov using the search term "telemedicine" and limiting results to randomized trials or observational studies registered between January 2012 and October 2023. We then reviewed trial registration data to determine if any of the outcomes assessed in the trials measured effect on clinicians' time.

Results: We found 113 studies and of these 78 studies of telemedicine met the inclusion criteria and were included. Nine (12 percent) of the 78 studies had some measure of clinician time as a primary outcome, and 11 (14 percent) as a secondary outcome. Four studies compared direct measures of TNT with telemedicine versus usual care, but no statistically significant difference was found. Of the sixteen studies including indirect measures of clinician time, thirteen found no significant effects, two found a statistically significant reduction, and one found a statistically significant increase.

Conclusions: This scoping review found that clinician time is not commonly measured in studies of telemedicine interventions. Attention to telemedicine's TNT in clinical studies and HTAs of telemedicine in practice may bring attention to the organization of clinical workflows and increase the value of telemedicine.

Alzheimer’s disease (AD) is a neurodegenerative disease with progressive neuropsychiatric symptoms. Patient care is often provided by informal caregivers similarly to various other chronic diseases. This targeted literature review assessed the difference in burden experienced by caregivers of people with AD in comparison to other chronic diseases.

Two separate search strings were developed to identify (i) caregiver burden in AD and (ii) caregiver burden in other chronic diseases using PubMed. Studies published in English (January 2012-October 2022) were included. Comparison of the caregiver burden was done using the weighted mean values (MV) of several questionnaires including the Zarit Burden Interview (ZBI), a 22-item self-report questionnaire for caregivers ranging from 0 to 88 points. ZBI is stratified into four categories of caregiving burden: Little or no burden (0 to 21), mild to moderate burden (22 to 40), moderate to severe burden (41 to 60) and severe burden (61 to 88).

ZBI was the most frequently used questionnaire; 13 studies reported data on caregiver burden in AD and 39 studies reported data on 20 other chronic diseases. The caregiver burden ranged from 18 to 48 in AD, measured by ZBI. The MV of AD burden was 36 based on a total of 1,703 participants. The caregiver burden in other chronic diseases ranged from MV of 5 (chronic musculoskeletal pain) to 59 (bipolar disorder).

Measured by ZBI, AD burden on caregivers (MV: 36, range: 18-48) was greater than heart failure (MV: 27, range: 16-29) and type 2 diabetes (MV: 26, range: Not reported) but lesser than schizophrenia (MV: 56, range: 52-65) and bipolar disorder (MV: 59, range: Not reported).

AD has a significant burden on caregivers. When assessing the value of interventions targeting AD, the impact of AD on caregivers should be considered in addition to the impact of AD on patients. Further studies are required to assess the informal care burden in AD and other chronic diseases.